I posted this question over at Linus Tech Tips a few days ago, but as there hasn't been any reply at all, I am trying it here now.
While there is a video where Linus Tech Tips explains CPU Boosts over here:
https://www.youtube.com/watch?v=H4ryOzIZvpQ
there aren't really any tests or guides about it. Most stuff is just about overclocking in general.
But I am wondering what you would get if you only overclick a couple of cpus via boost.
So let's say you get safe 4,6 GHz, but on 4,7 GHz the PC crashes. Now 4,6 GHz (or something in between) would be your max. However - what happens when you keep the CPU clock at a lower rate, let's say 4 GHz, but set the Boost for 2 CPUs up to 5 GHz?
Not sure about how to test it, especially it's stability. You probably would need a benchmark which could stress cores individually (haven't done any cpu benchmarking myself due to the lack of a proper rig). But I guess for games also something like this would be more efficient than having all cores on constant 4,6 GHz.
Also it would be interesting to see how processors with a different core count make differences. I mean can you achieve a higher boost on a 6 core cpu than on a 4 core cpu? How is hyperthreading affected by the boost?
Would be cool if someone has a couple of answers or if there is even a video about it.
While there is a video where Linus Tech Tips explains CPU Boosts over here:
https://www.youtube.com/watch?v=H4ryOzIZvpQ
there aren't really any tests or guides about it. Most stuff is just about overclocking in general.
But I am wondering what you would get if you only overclick a couple of cpus via boost.
So let's say you get safe 4,6 GHz, but on 4,7 GHz the PC crashes. Now 4,6 GHz (or something in between) would be your max. However - what happens when you keep the CPU clock at a lower rate, let's say 4 GHz, but set the Boost for 2 CPUs up to 5 GHz?
Not sure about how to test it, especially it's stability. You probably would need a benchmark which could stress cores individually (haven't done any cpu benchmarking myself due to the lack of a proper rig). But I guess for games also something like this would be more efficient than having all cores on constant 4,6 GHz.
Also it would be interesting to see how processors with a different core count make differences. I mean can you achieve a higher boost on a 6 core cpu than on a 4 core cpu? How is hyperthreading affected by the boost?
Would be cool if someone has a couple of answers or if there is even a video about it.